[Boost-testing] Re: Feature request: spamming the Boost list with the day's failures.

From: David Abrahams (dave_at_[hidden])
Date: 2005-06-02 12:53:12

Doug Gregor <dgregor_at_[hidden]> writes:

> On Jun 2, 2005, at 11:21 AM, David Abrahams wrote:
>> Pretty nice.  Here's the thing: compared to my original suggestion,
>> it's a bit hard for me to see if any of my libraries are failing.
>
> Part of your suggestion was for something like this:
>
>   Iterators:  12 regressions, 14 new warnings since release 1.32.0:
>      gcc-2.95-3: test1, test2, test3
>      msvc-8.0: test7, test9
>      ...
>
>   Graph: 3 regressions since release 1.32.0
>
>   Python:     245 regressions since release 1.32.0:
>      gcc-2.95-3: test1, test2, test3
>      msvc-8.0: test7, test9
>
> This would (I think) have the same problems as the issues list that's 
> available now. 

Maybe not if it didn't use tables.
Also, the appending of paren pairs to the test names seems to cause most
rows to be double-height in my browser.

> We could have a table of contents at the top of the 
> page, that links down to the set of failing tests for each library. 
> Then you could skim the table of contents to see if any of your 
> libraries are failing.

Sounds good.

>> One other point: when I click through any of the failures, I get "page
>> not found."
>
> You also see empty parens ( ) where you should see the name of the 
> regression test runner. 

Ah, than explains the parens.  I don't really need to see that
information up front.  How about putting it in the log text?

> Something is amiss and I'm looking into it as time permits. Click on
> the library name to see the full results for the library.

Thanks.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com